Obesity and the Economics of Prevention

7/27/2019 Obesity and the Economics of Prevention

1/270

Obesity and the Economicsof PreventionFIT NOT FAT

Franco Sassi


2/270


3/270

Obesityand the Economics

of Prevention

FIT NOT FAT

Franco Sassi


4/270

ORGANISATION FOR ECONOMIC CO-OPERATIONAND DEVELOPMENT

The OECD is a unique forum where governments work together to address the economic,

social and environmental challenges of globalisation. The OECD is also at the forefront of efforts

to understand and to help governments respond to new developments and concerns, such as

corporate governance, the information economy and the challenges of an ageing population.

The Organisation provides a setting where governments can compare policy experiences, seek

answers to common problems, identify good practice and work to co-ordinate domestic and

international policies.

The OECD member countries are: Australia, Austria, Belgium, Canada, Chile, the

Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy,

Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the

Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the

United States. The Commission of the European Communities takes part in the work of the OECD.

OECD Publishing disseminates widely the results of the Organisations statistics gathering

and research on economic, social and environmental issues, as well as the conventions,

guidelines and standards agreed by its members.

ISBN 978-92-64-06367-9 (print)ISBN 978-92-64-08486-5 (PDF)

Also available in French: Lobsit et lconomie de la prvention : Objectif sant

Photo credits: Cover Dmitriy Shironosov/Shuttlerstock.com Ju-Lee/Istockphoto.com Fotolia XI-Fotolia.com.

Corrigenda to OECD publications may be found on line at: www.oecd.org/publishing/corrigenda.

OECD 2010

You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications,databases and multimedia products in your own documents, presentations, blogs, websites and teaching materials, provided

that suitable acknowledgment of OECD as source and copyright owner is given. All requests for public or commercial use andtranslation rights should be submitted to [email protected]. Requests for permission to photocopy portions of this material forpublic or commercial use shall be addressed directly to the Copyright Clearance Center (CCC) at [email protected] or the Centrefranais dexploitation du droit de copie (CFC) at [email protected].

This work is published on the responsibility of the Secretary-General of the OECD. The opinions

expressed and arguments employed herein do not necessarily reflect the official views of the OECD or

of the governments of its member countries or those of the European Union.
http://www.oecd.org/publishing/corrigendahttp://www.oecd.org/publishing/corrigenda


5/270

FOREWORD

OBESITY AND THE ECONOMICS OF PREVENTION OECD 2010 3

Foreword

There was a fat boy in our street. People called him fatso, observes the maincharacter in Kieron Smith, Boy, a novel by James Kellman narrated from the point ofview of a child from the time he is 4 to almost 13. Through his eyes, we see a picture of

life in Glasgow in the 1960s and get an idea of the changes taking place. At the time,

obesity was unusual enough to draw attention. Yet now more than a third of Scottish 12-

year-olds are overweight, a fifth are obese and over one in ten severely obese. The

statistics for adults are even worse, with almost two-thirds of men and more than half

of women overweight. The situation is better in the other OECD countries, apart from the

United States, but obesity is a concern almost everywhere, in the OECD area and beyond.

If economics is the study of human behaviour as a relationship between ends

and scarce means which have alternative uses1 it must have something to say on

lifestyles, health and, above all, on the epidemic of obesity that has developed over the

past 30 years, one of the largest epidemics in the history of mankind. Indeed, obesity

has become a favourite subject for economists in various parts of the world, but the

role of economics in addressing the determinants and consequences of individual

health-related behaviours has been interpreted rather narrowly by many, including

some economists. This book is a humble attempt to explore the broader scope of the

potential contribution of economics to the design of effective, efficient and equitable

approaches to chronic disease prevention, with a focus on diseases linked to unhealthy

diets, sedentary lifestyles and obesity.

The public health paradigm, which still inspires and guides the field of chronic

disease prevention, is well reflected in Geoffrey Roses famous statement It is better to

be healthy than ill or dead. That is the beginning and the end of the only real argumentfor preventive medicine. It is sufficient.2 To an economist, Roses argument is of

critical importance, but it is not sufficient. And no sensible economist would claim that

what is missing is the economic argument that prevention will be a money-saver,

dismissed as misleading, or even false, by Rose himself. This book provides ample

evidence that Roses stance on this type of economic argument is well founded. The role

of economics is to ensure that prevention improves social welfare and its distribution

across social groups. This is what an economist would regard as a sufficient

argument for prevention. Health is one dimension of social welfare, but not the only one,

and not always the most important. Human behaviours are driven by many ends, touse Lionel Robbins word, which are all in competition with each other because resources

to pursue them are scarce. If so many people in the OECD area and beyond have been


6/270

FOREWORD

OBESITY AND THE ECONOMICS OF PREVENTION OECD 20104

gaining weight to the point that their health and longevity are affected, it may mean that

ends other than the pursuit of good health have taken a higher priority at a certain point

in time, or it may mean that peoples priorities have been increasingly constrained by

environmental influences, which they have not been able to handle. The role of economics

is to determine what mechanisms have been at play in the development of the obesity

epidemic and whether implementing actions that have the potential to reverse current

trends in obesity would generate an improvement in social welfare.

This book is the result of work undertaken at the OECD since 2007, following a

mandate received from the OECD Health Ministers at a meeting in Paris in 2004. The

book presents a wealth of data and analyses carried out by the OECD with the aim of

supporting the development of policies for tackling obesity and preventing chronic

diseases by its member countries. Some of these analyses were designed and

undertaken in close partnership with the World Health Organisation.

Notes

1. Lionel Robbins (1932), An Essay on the Nature and Significance of EconomicScience, Macmillan Facsimile, London.

2. Geoffrey Rose (1992), The Strategy of Preventive Medicine, Oxford University Press.


7/270

ACKNOWLEDGEMENTS


Acknowledgements

Many deserve credit for the contents of this book, but two deserve it aboveall: Michele Cecchini and Marion Devaux, whose tireless efforts have givensubstance to the work presented herein. Micheles work is behind the analysesof the impact of prevention strategies discussed in Chapter 6, while Marionsis behind all of the statistical analyses presented in Chapters 2and 3. Withoutthem, this book would not have been written. The author is also especiallygrateful to Jeremy Lauer and Dan Chisholm, who have made an invaluablecontribution to the assessment of the impact of prevention strategies andhave helped to establish, along with David Evans and Tessa Tan-Torres, a mostproductive collaboration between the OECD and the WHO on the economics ofchronic disease prevention. Other OECD colleagues who provided valuablecontributions to the work at various stages of the Economics of Preventionproject include Jeremy Hurst, Linda Fulponi, Mark Pearson, Peter Scherer,El izabeth Docteur, John Mart in, Mart ine Durand, Elena Rusticel li ,

Christine Le Thi and Francesca Borgonovi, as well as Anna Ceccarelli,Jody Church, Amrita Palr iwala , J i Hee Yo un, Fareen Hassan,Romain Lafarguette, Angelica Carletto and Lucia Scopelliti who worked on theEconomics of Prevention project during internships in the OECD HealthDivision. Members of the Expert Group on the economics of preventionnominated by OECD countries, too many to list individually, as well asmembers invited by the OECD Secretariat, including Donald Kenkel,Marc Suhrcke, Evi Hatziandreu, Edward Glaeser, Francesco Branca,Thomas Philipson, Tim Lobstein, Klim McPherson, Julia Critchley, Taavi Lai,

Godfrey Xuereb, and Mike Murphy have greatly improved the quality of thework presented in this book. Several of them have contributed directly to thebook, in the special focus sections which follow some of the chapters. Theauthor is also grateful to representatives of the food and beverage industryand of the sports and exercise industry who provided comments on projectplans and outputs through the Business and Industry Advisory Committee tothe OECD (BIAC). Country analyses of the impact of prevention strategies weremade possible by inputs received from Sylvie Desjardins, Jacques Duciaumeand Peter Walsh (Canada), Peter Dick and Francis Dickinson (England),

Giovanni Nicoletti and Stefania Vasselli (Italy), Nobuyuki Takakura,Kaori Nakayama, Shunsaku Mitzushima, Tetsuya Fijikawa and Hitoshi Fujii


8/270

ACKNOWLEDGEMENTS


(Japan), Fernando Alvarez Del Rio, Cristina Gutierrez Delgado, Gustavo RiveraPena and Veronica Guajardo Barron (Mexico), who also helped to interpret thefindings of the analyses. Finally, the author acknowledges the continuedsupport, encouragement and helpful comments received from the OECD

Health Committee, chaired by Jane Halton, throughout the duration of theEconomics of Prevention project.

Special thanks go to Tracey Strange and Marlne Mohier for their mostvaluable editorial contributions, to Patrick Love for contributions at anearlier stage in the development of the book, and to Kate Lancaster andCatherine Candea for their help in transforming an editorial project into a realpublication. Further editorial assistance was provided during the course of theproject by Gabrielle Luthy, Christine Charlemagne, Elma Lopes, Aidan Curran,Judy Zinnemann and Isabelle Vallard.

The Economics of Prevention project was partly funded through regularcontributions from OECD member countries. Additional voluntarycontributions to the project were made by the following member countries:Australia, Canada, Denmark, Italy, Japan, Mexico, Netherlands, Sweden,Switzerland and United Kingdom. The project was also partly supported by agrant from the Directorate General for Public Health and Consumer Affairs ofthe European Commission. The contents of this book do not necessarily reflectthe views of the Commission.


9/270

TABLE OF CONTENTS


Table of ContentsAbbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Chapter 1. Introduction: Obesity and the Economics of Prevention. . . . . 23Obesity: The extent of the problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Obesity, health and longevity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26The economic costs of obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28The implications for social welfare and the role of prevention . . . . . 30What economic analyses can contribute . . . . . . . . . . . . . . . . . . . . . . . . 32The books main conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Overview of the remaining chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Key messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

Special Focus I. Promoting Health and Fighting Chronic Diseases:

What Impact on the Economy? (by Marc Suhrcke) . . . . . . . 49Chapter 2. Obesity: Past and Projected Future Trends . . . . . . . . . . . . . . . . 57

Obesity in the OECD and beyond . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Measuring obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59Historical trends in height, weight and obesity. . . . . . . . . . . . . . . . . . . 61Cohort patterns in overweight and obesity . . . . . . . . . . . . . . . . . . . . . . 65Projections of obesity rates up to 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . 67Key messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76Chapter 3. The Social Dimensions of Obesity . . . . . . . . . . . . . . . . . . . . . . . . 79

Obesity in different social groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80Obesity in men and women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80Obesity at different ages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Obesity and socio-economic condition. . . . . . . . . . . . . . . . . . . . . . . . . . 82Obesity in different racial and ethnic groups. . . . . . . . . . . . . . . . . . . . . 95Does obesity affect employment, wages and productivity? . . . . . . . . 97Key messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103


10/270

TABLE OF CONTENTS


Special Focus II. The Size and Risks of the International Epidemicof Child Obesity (by Tim Lobstein) . . . . . . . . . . . . . . . . . . . 107

Chapter 4. How Does Obesity Spread? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

The determinants of health and disease . . . . . . . . . . . . . . . . . . . . . . . . 116The main driving forces behind the epidemic. . . . . . . . . . . . . . . . . . . . 121Market failures in lifestyle choices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122The social multiplier effect: Clustering of obesity withinhouseholds, peer groups and social networks. . . . . . . . . . . . . . . . . . . . 129Key messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Special Focus III. Are Health Behaviors Driven by Information?(by Donald Kenkel). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Chapter 5. Tackling Obesity: The Roles of Governments and Markets . . 147What can governments do to improve the qualityof our choices? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148Government policies on diet and physical activityin the OECD area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154Private sector responses: Are markets adjustingto the new challenges? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158Key messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162

Special Focus IV. Community Interventions for the Preventionof Obesity (by Francesco Branca) . . . . . . . . . . . . . . . . . . . . . 165

Chapter 6. The Impact of Interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175What interventions really work? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176Cost-effectiveness analysis: A generalised approach . . . . . . . . . . . . . . 186Effects of the interventions on obesity, healthand life expectancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189The costs and cost-effectiveness of interventions . . . . . . . . . . . . . . . . 194Strategies involving multiple interventions. . . . . . . . . . . . . . . . . . . . . . 198

Distributional impacts of preventive interventions . . . . . . . . . . . . . . . 201From modelling to policy: Key drivers of success . . . . . . . . . . . . . . . . . 203Key messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

Special Focus V. Regulation of Food Advertising to Children:the UK Experience (by Jonathan Porter) . . . . . . . . . . . . . . . 211

Special Focus VI. The Case for Self-Regulation in FoodAdvertising (by Stephan Loerke) . . . . . . . . . . . . . . . . . . . . . 217


11/270

TABLE OF CONTENTS


Chapter 7. Information, Incentives and Choice:A Viable Approach to Preventing Obesity . . . . . . . . . . . . . . . . . 221

Tackling the obesity problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222Populations or individuals? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

Changing social norms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225A multi-stakeholder approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228How much individual choice? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232Key messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235

Annex A. Supplementary Figures and Tables . . . . . . . . . . . . . . . . . . . . . . . . 237

Annex B. Authors and Contributors Biographies . . . . . . . . . . . . . . . . . . . . 263

Tables

SFII.1. Estimated prevalence of excess body weight in school-agechildren in 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

A.1. Description of the national health survey data usedin the analyses reported in Chapters 2 and 3 . . . . . . . . . . . . . . . . . 238

A.2. Main input parameters used in CDP model-based analysesand relevant sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243

A.3. Costs and coverage of selected preventive interventions . . . . . . . 249A.4. Magnitude of health gains associated with preventive

interventions (population per DALY/LY gained) . . . . . . . . . . . . . . . 250

Figures

2.1. Obesity and overweight in OECD and non-OECD countries . . . . . . 602.2. Age-standardised obesity rates, age 15-64,

selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632.3. Age-standardised overweight rates, age 15-64,

selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642.4. Cohort patterns in obesity and overweight

in selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672.5. Changes over time in the BMI distribution in Australia

and England . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692.6. Past and projected future rates of obesity and overweight,

age 15-74, selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . 702.7. Past and projected future rates of child obesity and

overweight, age 3-17, in four OECD countries. . . . . . . . . . . . . . . . . 733.1. Obesity and overweight by age in six OECD countries . . . . . . . . . 823.2. Obesity by education level in four OECD countries . . . . . . . . . . . . 853.3. Years spent in full-time education according to obesity

status at age 20, France, population aged 25-65 . . . . . . . . . . . . . . . 87

3.4. Disparities in obesity and overweight by education level,selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93


12/270

TABLE OF CONTENTS


3.5. Disparities in obesity and overweight by household incomeor occupation-based social class, selected OECD countries . . . . . 94

3.6. Obesity and overweight by ethnic group in England (adults) . . . . 953.7. Obesity and overweight by ethnic group in the United States

(adults) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963.8. Obesity and overweight by ethnic group in England

(children 3-17) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973.9. Obesity and overweight by ethnic group in the United States

(children 3-17) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97SFII.1. Estimated prevalence of child overweight in OECD member

states and associated countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108SFII.2. Trends in prevalence of overweight among children in England,

France and United States (obese only) . . . . . . . . . . . . . . . . . . . . . . 110

4.1. Child obesity and overweight by parents obesity status . . . . . . . 1324.2. BMI correlation between spouses and between mothers

and children. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1334.3. BMI Correlation in couples of different ages. . . . . . . . . . . . . . . . . . 1335.1. Interventions in OECD and other EU countries by type. . . . . . . . . 1565.2. Interventions in OECD and other EU countries by sector . . . . . . . 1576.1. Health outcomes at the population level

(average effects per year) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1916.2. Cumulative DALYs saved over time . . . . . . . . . . . . . . . . . . . . . . . . . 192

6.3. Effects of selected interventions in different age groups . . . . . . . 1936.4. Cumulative impact on health expenditure over time . . . . . . . . . . 1946.5. Economic impact at the population level

(average effects per year) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1956.6. Cost-effectiveness of interventions over time . . . . . . . . . . . . . . . . 1976.7. Estimated impacts of a multiple-intervention strategy

(average effects per year) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200SFVI.1. A blueprint for marketing policies on food advertising. . . . . . . . . 218

A.1. Obesity by household income or occupation-based social class,selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

A.2. Overweight by household income or occupation-based socialclass, selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240

A.3. Obesity by education level, selected OECD countries . . . . . . . . . . 241A.4. Overweight by education level, selected OECD countries . . . . . . . 242A.5. Cumulative DALYs saved over time (per million population). . . . 251A.6. Cumulative impact on health expenditure over time . . . . . . . . . . 253A.7. Cumulative DALYs saved with a multiple-intervention

strategy over time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255A.8. Cumulative impact on health expenditure

of a multiple-intervention strategy over time. . . . . . . . . . . . . . . . . 255A.9. Cost-effectiveness of a multiple-intervention strategy over time . 256


13/270

TABLE OF CONTENTS


A.10. Canada: Probabilistic sensitivity analysis of the cost-effectivenessof interventions at 30 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

A.11. England: Probabilistic sensitivity analysis of the cost-effectivenessof interventions at 30 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

A.12. Italy: Probabilistic sensitivity analysis of the cost-effectivenessof interventions at 30 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258

A.13. Japan: Probabilistic sensitivity analysis of the cost-effectivenessof interventions at 30 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258

A.14. Mexico: Probabilistic sensitivity analysis of the cost-effectivenessof interventions at 30 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

A.15. Canada: Probabilistic sensitivity analysis of the cost-effectivenessof interventions at 100 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

A.16. England: Probabilistic sensitivity analysis of the cost-effectiveness

of interventions at 100 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260A.17. Italy: Probabilistic sensitivity analysis of the cost-effectiveness

of interventions at 100 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260A.18. Japan: Probabilistic sensitivity analysis of the cost-effectiveness

of interventions at 100 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261A.19. Mexico: Probabilistic sensitivity analysis of the cost-effectiveness

of interventions at 100 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261

This book has...

StatLinks2A service that delivers Excel files

from the printed page!

Look or the StatLinks at the bottom right-hand corner o the tables or graphs in this book.

To download the matching Excel spreadsheet, just type the link into your Internet browser,

starting with thehttp://dx.doi.org prefx.

I youre reading the PDF e-book edition, and your PC is connected to the Internet, simply

click on the link. Youll fnd StatLinks appearing in more OECD books.


14/270


15/270

ABBREVIATIONS


Abbreviations

ADLs Activities of daily livingAPC Age-period-cohortANGELO Analysis grid for environments linked to obesityBMI Body mass index

CDP Chronic Disease PreventionCEA Cost-effectiveness analysisCHD Coronary heart diseaseCHOICE Choosing interventions that are cost effective (WHO project)COI Cost of illnessCONAPO Consejo Nacional de Poblacin (Mexico)DALY Disability-adjusted life yearDGIS Direccin General de Informacin en Salud (Mexico)EPODE Ensemble, Prvenons lObsit des Enfants (European Project)FSA Food Standards Agency (United Kingdom)GBP Great Britain PoundGCEA Generalised cost-effectiveness analysisGDP Gross domestic productGEMS Girls Health Enrichment Multi-site Studies (Stanford)HBSC Health behaviour in school-aged childrenHFSS High in fat, salt and sugarHSE Health Survey for EnglandIARC International Agency for Research on CancerIFBA International Food and Beverage AllianceIOM Institute of Medicine (United States)IMSS Instituto Mexicano del Seguro Social (Mexico)ISTAT Istituto Nazionale di Statistica (Italy)LY Life yearMCBS Medicare Current Beneficiary Survey (United States)MoH Ministry of HealthNBGH National Business Group on Health (United States)NGO Non-governmental organisationNHANES National Health and Nutrition Examination Survey (United States)NHIS National Health Interview Survey (United States)NIPH National Institute of Public Health (Japan)


16/270

ABBREVIATIONS


NLEA Nutrition Labelling and Education Act (United States)OPIC Obesity Prevention In CommunitiesPHAC Public Health Agency of CanadaPPPs Purchasing power paritiesQALY Quality-adjusted life yearRR Relative riskRRa Relative rateSES Socio-economic statusUSD American dollarUSDA US Department of AgricultureWFA World Federation of AdvertisersWHO World Health Organisation


17/270

Obesity and the Economics of Prevention

Fit not Fat OECD 2010

15

Executive Summary

Obesity is a major health concern for OECD countries. Using a wide range ofindividual-level and population data from OECD countries, this book presentsanalyses of trends in obesity, explores the complex causal factors affecting theepidemic and develops an assessment of the impacts interventions to combatthe problem. The book provides new information on what prevention

strategies are most effective and cost-effective, discussing the respective rolesof individuals, social groups, industry and government, and the implicationsof these findings for the development of policies to address one of the largestpublic health emergencies of our time.

The book presents an economic approach to the prevention of chronicdiseases, which recognises the importance of human goals that arepotentially in competition with the pursuit of good health and the social andmaterial constraints which influence individual choice and behaviours. Aneconomic approach aims at identifying possible factors, technically marketfailures, which limit opportunities for people to make healthy lifestylechoices, and devising suitable strategies to overcome such failures.

What are the health and economic costs associated with obesity?

Chapter 1 places obesity in the contextof the growing burden of chronic diseaseand discusses the extent of the problem

Much of the burden of chronic diseases is linked to lifestyles, with tobaccosmoking, obesity, diet and lack of physical activity being responsible for thelargest shares of such burden. Research has shown that people who lead aphysically active life, do not smoke, drink alcohol in moderate quantities, and eatplenty of fruits and vegetables have a risk of death that is less than one fourth ofthe risk of those who have invariably unhealthy habits. Mortality increasessteeply once individuals cross the overweight threshold. The lifespan of an obeseperson is up to 8-10 years shorter (for a BMI of 40-45) than that of a normal-weightperson, mirroring the loss of life expectancy suffered by smokers. An overweight

person of average height will increase their risk of death by approximately 30% forevery 15 additional kilograms of weight. In ten European countries, the odds of


18/270

EXECUTIVE SUMMARY


disability, defined as a limitation in activities of daily living (ADL), are nearly twiceas large among the obese as in normal weight persons.

An obese person generates higher health care expenditures than a normal-

weight person and costs increase disproportionately at increasing levels ofBMI. However, over a lifetime, existing estimates suggest that an obese persongenerates lower expenditures than a person of normal weight (but higher thana smoker, on average).

What are the trends in obesity past and future?

Chapter 2 looks at the developmentof obesity over time and its relationshipto changes in diet and lifestyle

Height and weight have been increasing since the 18th century in many of thecurrent OECD countries, as income, education and living conditions graduallyimproved over time. Surveys began to record a sharp acceleration in the rate ofincrease in body mass index (BMI) in the 1980s, which in many countries grewtwo to three times more rapidly than in the previous century. While gains inBMI had been largely beneficial to the health and longevity of our ancestors,an alarming number of people have now crossed the line beyond whichfurther gains become more and more detrimental. Before 1980, obesity rates

were generally well below 10%. Since then, rates have doubled or tripled inmany countries, and in almost half of OECD countries 50% or more of thepopulation is overweight.

Rates of overweight and obesity vary considerably across OECD countries, buthave been increasing consistently over the past three decades everywhere. Ifrecent trends in OECD countries continue over the next ten years, projectionssuggest that pre-obesity rates (a BMI above the normal limit of 25 but below theobesity level of 30) for the 15-74 age group will stabilise progressively, and mayeven shrink slightly in many countries, while obesity rates continue to rise.

On the one hand, obesogenic environments, including physical, social andeconomic environments, have contributed to higher obesity rates over thepast 30 years by exerting powerful influences on peoples overall calorieintake, on the composition of their diets and on the frequency and intensity ofphysical activity at work, at home and during leisure time. On the other hand,changing individual attitudes, reflecting the long-term influences of improvededucation and socio-economic conditions, have countered environmentalinfluences to some extent.


19/270

EXECUTIVE SUMMARY


Which groups are the most affected by obesity?What are the social impacts of obesity?

Chapter 3 looks at how age, gender

education and socio-economic status affectobesity rates and, conversely, at how obesityaffects labour market opportunitiesand outcomes

There does not appear to be a uniform gender pattern in obesity acrosscountries. Worldwide, obesity rates tend to be higher in women than in men,other things being equal, and the same is true, on average, in the OECD area.Male obesity rates have also been growing faster than female rates in most

OECD countries. The gender dimension is especially important because of itssignificant interactions with other individual characteristics, such as socio-economic condition or ethnicity.

A complex relationship exists between socio-economic condition and obesity.For example, this relationship changes as economies become more developed,with poorer people more likely to be affected in rich countries. Analyses of datafrom more than one third of OECD countries show important social disparitiesin overweight and obesity in women and lesser or no disparities in men. Socialdisparities within countries are larger in obesity than in overweight, but when

comparisons across countries are made, the size of disparities is not related tocountries overall obesity rates. With few exceptions, social disparities inobesity remained remarkably stable over the past 15 years.

Social disparities are also present in children in three out of four countriesexamined, but no major differences between genders are observed in degreesof disparity. The gap in obesity between children who belong to ethnicminorities and white children in England and in the United States is largerthan that observed in adults.

Disparities in labour market outcomes between the obese and people of

normal weight, which are particularly strong in women, are likely tocontribute to the social gradient in overweight and obesity. The obese are lesslikely to be part of the labour force and to be in employment. Discriminationin hiring decisions, partly due to expectations of lower productivity,contributes to the employment gap. White women are especiallydisadvantaged in this respect. The obese are likely to earn less than people ofnormal weight. Wage penalties of up to 18% have been associated with obesityin existing research. The obese tend to have more days of absence from work,a lower productivity on the job and a greater access to disability benefits than

people of normal weight. The need for government intervention to protect theobese in labour markets and ensure they enjoy the same opportunities as


20/270

EXECUTIVE SUMMARY


anyone else in terms of employment, type of job, sector of occupation and paynaturally follows the evidence presented in Chapter 3.

How did obesity become a problem?Chapter 4 explores some of the key dynamicsthat have contributed to the obesityepidemic, including the role of factors whichhave made it difficult for individuals tohandle increasing environmental pressures

The obesity epidemic is the result of multiple, complex and interactingdynamics, which have progressively converged to produce lasting changes in

peoples lifestyles. The supply and availability of food have changedremarkably in the second half of the 20th century, in line with major changesin food production technologies and an increasing and increasinglysophisticated use of promotion and persuasion. The price of calories felldramatically and convenience foods became available virtually everywhere atany time, while the time available for traditional meal preparation from rawingredients shrunk progressively as a result of changing working and livingconditions. Decreased physical activity at work, increased participation ofwomen in the labour force, increasing levels of stress and job insecurity,

longer working hours for some jobs, are all factors that, directly or indirectly,contributed to the lifestyle changes which caused the obesity epidemic.

Government policies have also played a part in the obesity epidemic.Examples include subsidies (e.g. in agriculture) and taxation affecting theprices of lifestyle commodities; transport policies, some of which have led toan increased use of private means of transportation; urban planning policiesleaving scarce opportunities for physical activity, or leading to the creation ofdeprived and segregated urban areas that provide fertile grounds for thespread of unhealthy lifestyles and ill health.

The question must be asked of whether the changes that fuelled obesity andchronic diseases in the past decades are simply the outcome of efficientmarket dynamics, or the effect of market and rationality failures preventingindividuals from achieving more desirable outcomes. In the design andimplementation of prevention policies special attention must be placed on therole of information, externalities and self-control issues, including the role ofsocial multiplier effects (the clustering and spread of overweight and obesitywithin households and social networks) in the obesity epidemic. Evidence ofsimilar failures is reviewed and the scope for prevention to address some of

the consequences of those failures is discussed in the book.


21/270

EXECUTIVE SUMMARY


What can governments and markets do to improve health-relatedbehaviours?

Chapter 5 looks at the broad range of actions

taken in recent years to improve nutritionand physical activity in OECD countries

Governments can increase choice by making new healthy options available, orby making existing ones more accessible and affordable. Alternatively, they canuse persuasion, education and information to make healthy options moreattractive. These are often advocated as minimally intrusive interventions, butgovernments may not always deliver persuasion effectively and in the bestinterest of individuals, and it is difficult to monitor whether they do so.

Regulation and fiscal measures are more transparent and contestableinterventions, although they hit all consumers indiscriminately, may be difficultto organise and enforce and may have regressive effects. Interventions that areless intrusive on individual choices tend to be more expensive, whileinterventions that are more intrusive have higher political and welfare costs.

A survey of national policies in 2007-08 covering all OECD and EU countriesshows that governments acknowledge that individuals are often exposed tolarge amounts of potentially confusing information on health and lifestylesfrom a variety of sources, and assert that it is primarily their responsibility to

act as a balanced and authoritative source of information, thus providing clearguidance to individuals who struggle to cope with increasingly powerfulenvironmental influences.

Many governments are intensifying their efforts to promote a culture ofhealthy eating and active living. A large majority of them have adoptedinitiatives aimed at school-age children, including changes in the schoolenvironment, notably regarding food and drink, as well as improvements infacilities for physical activity. The second most common group ofinterventions involves the public health function of health systems. These

interventions are primarily based on the development and dissemination ofnutrition guidelines and health promotion messages to a wide variety ofpopulation groups through numerous channels, as well as promotion of activetransport and active leisure. Governments have been more reluctant to useregulation and fiscal levers because of the complexity of the regulatoryprocess, the enforcement costs involved, and the likelihood of sparking aconfrontation with key industries.

The private sector, including employers, the food and beverage industry, thepharmaceutical industry, the sports industry and others, has made a

potentially important contribution to tackling unhealthy diets and sedentarylifestyles, often in co-operation with governments and international


22/270

EXECUTIVE SUMMARY


organisations. Evidence of the effectiveness of private sector interventions isstill insufficient, but an active collaboration between the public and theprivate sector will enhance the impact of any prevention strategies and spreadthe costs involved more widely. Key areas in which governments expect a

contribution from the food and beverage industry are: food productreformulation; limitation of marketing activities, particularly to vulnerablegroups; transparency and information about food contents.

What interventions work best and at what cost?

Chapter 6 presents a comprehensiveanalysis of the impacts of nine differenthealth interventions on obesity and related

chronic diseases in five OECD countries:Canada, England, Italy, Japan and Mexico

Interventions aimed at tackling obesity by improving diets and increasingphysical activity in at least three areas, including health education andpromotion, regulation and fiscal measures, and counselling in primary care, areall effective in improving health and longevity and have favourable cost-effectiveness ratios relative to a scenario in which chronic diseases are treatedonly as they emerge. When interventions are combined in a multiple-

intervention strategy, targeting different age groups and determinants ofobesity simultaneously, overall health gains are significantly enhanced withoutany loss in cost-effectiveness. The cost of delivering a package of interventionswould vary between USD PPPs 12 per capita in Japan to USD PPPs 24 in Canada,a tiny fraction of health expenditure in those countries, and also a smallproportion of what is currently spent on prevention in the same countries.

Most of the interventions examined have the potential to generate gainsof 40 000 to 140 000 years of life free of disability in the five countries together,with one intervention, intensive counselling of individuals at risk in primary

care, leading to a gain of up to half million life years free of disability. However,counselling in primary care is also the most expensive of the interventionsconsidered in the analysis. Interventions with the most favourable cost-effectiveness profiles are outside the health care sector, particularly in theregulatory and fiscal domain. Interventions, especially those aimed atchildren, may take a long time to make an impact on peoples health and reachfavourable cost-effectiveness ratios.

Interventions add years of healthy life to peoples health expectancy, reducinghealth care costs. However, the health benefits of prevention are such that

people also live longer with chronic diseases, and years of life are added in theoldest age groups, increasing the need for health care. The interventions


23/270

EXECUTIVE SUMMARY


assessed may, at best, generate reductions in the order of 1% of totalexpenditure for major chronic diseases. At the same time, many suchinterventions involve costs which outweigh any reductions in healthexpenditure. These costs may arise in different jurisdictions. Some are typically

paid through public expenditure, but do not necessarily fall within health carebudgets (e.g. the costs associated with regulatory measures, or interventions onthe education or transport systems). Others fall outside public budgetsaltogether (e.g. most of the costs associated with worksite interventions).

The distributional impacts of interventions are mostly determined bydifferences in morbidity and mortality among socio-economic groups. Fiscalmeasures are the only intervention producing consistently larger health gainsin the less well-off. The distributional impacts of other interventions vary indifferent countries.

Those reported in Chapter 6 are likely to be conservative estimates of theimpacts to be expected in real world settings. Key drivers of success forpreventive interventions include high participation (on both supply anddemand sides), long-term sustainability of effects, ability to generate socialmultiplier effects, and combination of multiple interventions producing theireffects over different time horizons.

How can an unhealthy societal trend be turned around?

Chapter 7 outlines the role of information,incentives and choice in designing policiesto combat obesity and discusses therelevance of a multi-stakeholder approachto chronic disease prevention

The main question addressed in this book is how to trigger meaningfulchanges in obesity trends. The short answer is by wide-ranging preventionstrategies addressing multiple determinants of health. The reality is that

every step of the process is conditioned not just by public health concerns, butby history, culture, the economic situation, political factors, social inertia andenthusiasm, and the particularities of the groups targeted.

Individual interventions have a relatively limited impact; therefore,comprehensive strategies involving multiple interventions to address a rangeof determinants are required to reach a critical mass one that can have ameaningful impact on the obesity epidemic by generating fundamentalchanges in social norms. The development of comprehensive preventionstrategies against obesity needs to focus on how social norms are defined andhow they change; on the influence of education and information on obesitybut also on the potential for government regulation to affect behaviours; and


24/270

EXECUTIVE SUMMARY


on the role of individual choice and values. A sensible prevention strategyagainst obesity would combine population and individual (high-risk)approaches, as the two have different and complementary strengths in thepursuit of effectiveness, efficiency and favourable distributional outcomes.

The adoption of a multi-stakeholder approach is increasingly invoked as themost sensible way forward in the prevention of chronic diseases. But whilefew if any of those involved would argue with this in theory, the interests ofdifferent groups are sometimes in conflict with each other and it is not alwayspossible to find a solution where nobody loses out. Yet at the same time, noparty is in a position to meaningfully reduce the obesity problem andassociated chronic diseases without full co-operation with other stakeholders.


25/270

Obesity and the Economics of Prevention

Fit not Fat OECD 2010

23

Chapter 1

Introduction: Obesityand the Economics of Prevention

Unprecedented improvements in population health have beenrecorded in OECD countries during the past century, thanks toeconomic growth and to public policies in education, sanitation,health, and welfare. Yet industrialisation and prosperity have beenaccompanied by increases in the incidence of a number of chronic

diseases, for which obesity is a major risk factor. This chapter looksat the impact of obesity on health and longevity and the economiccosts that obesity generates, now and for the future. It examinesthe role of prevention in mitigating these effects and presents acase for how an economic perspective on the prevention of chronicdiseases linked to lifestyles and obesity can provide insight intobetter ways of addressing the obesity epidemic.


26/270

1. INTRODUCTION: OBESITY AND THE ECONOMICS OF PREVENTION


Obesity: The extent of the problem

Unprecedented improvements in population health have been recordedin OECD countries during the past century. Life expectancy has increased onaverage by as much as 25-30 years. Major infectious diseases have beeneradicated. Infant mortality rates have been dramatically reduced. Peoplehave gained in height and weight over time, with a substantial numbermoving out of under-nutrition. Economic growth has played an important role

in these achievements, and so have public policies in education, sanitation,public health, and the development of welfare systems. However,industrialisation and prosperity have been accompanied by increases in theincidence of a number of chronic diseases. Advances in medical care have, insome cases, prevented increasing incidence from translating into highermortality, but industrialised societies bear growing burdens of disability,which are contributing to rising health care expenditures.

Lifestyles have played an important part in the health changes describedhere. In high-income countries, smoking alone is estimated to be responsible for

22% of cardiovascular diseases, and for the vast majority of some cancers andchronic respiratory diseases. Alcohol abuse is deemed to be the source of 8-18% ofthe total burden of disease in men and 2-4% in women. Overweight and obesityaccount for an estimated 8-15% of the burden of disease in industrialisedcountries, while high cholesterol accounts for 5-12% (WHO, 2002).

Studies conducted in the 1970s and 1980s in the county of Alameda,California, showed that healthy habits concerning aspects of diet, physicalactivity, smoking, alcohol consumption and sleeping patterns could reducemortality rates by 72% in men and 57% in women, relative to rates observed inthose who had mostly unhealthy habits (Breslow and Enstrom, 1980). A recentstudy in England produced similar findings, suggesting that combininghealthy habits has the strongest impact on mortality. People who lead aphysically active life, do not smoke, drink alcohol in moderate quantities, andeat plenty of fruits and vegetables have a risk of death that is less than onefourth of the risk of those who have invariably unhealthy habits (Khaw et al.,2008). In Ireland, almost half of the reduction in CHD mortality ratesduring 1985-2000 in the age group 25-84 was attributed to declining trends inthe number of smokers and in the mean levels of cholesterol and bloodpressure (Bennet et al., 2006). Active lifestyle change may reap large benefits,as demonstrated, for instance, by a 25-year intervention on adult men in


27/270



Finland, named the North Karelia project, which is purported to have led to a68% decline in cardiovascular disease mortality, 73% in coronary heart disease,44% in cancer, 71% in lung cancer, and to a 49% decline in deaths from allcauses (Puska et al., 1998).

Among the many epidemics that hit the world in the 20th century, twohave contributed to a substantial proportion of the burden of chronic diseases,especially in high-income countries: tobacco smoking and obesity.

Cigarette smoking was a phenomenon of negligible importance in theearly 1900s, but smoking rates increased steadily during the course of thecentury, in line with the mass production of cigarettes. The increase wasparticularly large between the 1930s and the 1960s. During the 1960sand 1970s, smoking rates reached peaks of 50% or more in many OECD

countries, before starting to decline.Solid evidence of the harm caused by tobacco to the health of smokers

has been available at least since the 1950s. In 1964, the US Surgeon Generalissued a landmark report outlining the sheer scope of the health risksassociated with smoking. However, it took many more years for the addictivenature of tobacco and the dangers of passive smoking to be fully and widelyrecognised, amidst deceptive actions by the tobacco industry and a heavyinvolvement of the judiciary.

The obesity epidemic has developed more recently. Height and weight

have been increasing since the 18th century in many of the current OECDcountries, as income, education and living conditions gradually improved overtime. Surveys began to record a sharp acceleration in the rate of increase inbody mass index (BMI) in the 1980s, which in many countries grew two tothree times more rapidly than in the previous century. While gains in BMI hadbeen largely beneficial to the health and longevity of our ancestors, analarmingly large number of people have now crossed the line beyond whichfurther gains become more and more detrimental. Before 1980, obesity rateswere generally well below 10%. Since then, rates have doubled or tripled in

many countries, and in almost half of OECD countries 50% or more of thepopulation is overweight.

Evidence of a link between body weight and mortality dates back to theearly 1950s (Dublin, 1953), but the harmful effects of specific nutrients andthose of increasingly sedentary jobs and lives has proved much more difficultto ascertain. It was only in recent years that a clear link between unsaturated(trans) fats, particularly hydrogenated oils, and coronary heart disease wasestablished (Mozaffarian and Stampfer, 2010). But for most nutrients,including other types of fats, sugar and salt, the issue is rather to determine at

what levels their consumption may become a health hazard. The factors thatinfluence what people eat and the activities in which they engage are so many


28/270



and so diverse that capturing the fundamental causes of the obesity epidemicand acting on the levers which may effectively and durably change the courseof the epidemic is a considerable challenge.

Obesity, health and longevity

Obesity is a major public health concern because it is a key risk factor fora range of chronic diseases (Malnick and Knobler, 2006), with diabetes beingthe most closely linked. The severely obese have a risk of developing type 2diabetes up to 60 times larger than those at the lower end of the normalweight spectrum. High blood pressure and high cholesterol are also morecommon as BMI increases. These links make the obese more likely to developheart disease, particularly coronary artery disease, and stroke, and to die from

these diseases. A large proportion of major cancers such as breast andcolorectal cancer is linked to obesity and physical inactivity. Obesity alsoincreases the chances of developing a number of respiratory andgastrointestinal diseases, as well as osteoarthritis, some mental conditions,and many other diseases and complaints, too numerous to list here. Some ofthe consequences of obesity may not even be known yet.

Chronic diseases are currently the main cause of both disability anddeath worldwide. They affect people of all ages and social classes, althoughthey are more common in older ages and among the socially disadvantaged

(WHO, 2002). Globally, of the 58 million deaths that occurred in 2005,approximately 35 million, or 60%, were due to chronic causes. Most deathswere due to cardiovascular disorders and diabetes (32%), cancers (13%), andchronic respiratory diseases (7%) (Abegunde et al., 2007). This burden ispredicted to worsen in the coming years. A WHO study projected an increaseof global deaths by a further 17% in the period 2005-15, meaning that of the64 million estimated deaths in 2015, 41 million people will die of a chronicdisease (WHO, 2005).

The burden of chronic diseases is proportionally even larger in OECD

countries. In 2002, these caused 86% of deaths in the European region (WHO,2004). However, the prevalence of many chronic diseases, including diseases ofthe circulatory system, digestive and respiratory diseases, was substantiallylower at the end of the 20th century than it had been at the start of the centuryin countries such as the United States (Fogel, 1994). Mortality forcardiovascular diseases more than halved in the United States in the latterpart of the last century, after the end of World War II. Deaths decreased by afurther 13% between 1996 and 2006, as case fatality dropped by almost 30%. Inmany countries, mortality declined more rapidly among the better off. Social

disparities in premature mortality from cardiovascular diseases and manycancers widened in countries such as Finland, Norway, Denmark, Belgium,Austria and England (Mackenbach, 2006).


29/270



Such a dramatic fall in mortality, which was not mirrored by comparabledeclines in disease incidence, and a general increase in longevity, led to asubstantial growth of morbidity associated with chronic diseases in recentyears. In Denmark, an estimated 40% of the population lives with long-term

conditions (WHO Europe, 2006), while in the United States the majority of70-year-olds is affected by at least one chronic condition, with cardiovasculardiseases alone affecting 40% of males (Adams et al., 1999). OECD researchshowed a generalised increase in the prevalence of diabetes among theelderly. Alarming trends were observed even in countries traditionallyminimally affected by such disease. For instance, Japan saw a 5.3% averageannual increase in the prevalence of diabetes in the period 1989-2004(Lafortune and Balestat, 2007). Co-morbidities also increase with age, andpopulations are ageing rapidly in the OECD area. In western Europe, the

number of people aged over 64 has more than doubled in the last 60 years,while the number of those aged over 80 has quadrupled. As a consequence,several chronic diseases can co-exist in many individuals. At least 35% of menover 60 years of age have been found to have two or more chronic conditions(WHO Europe, 2006), and of the 17 million people living with long-termchronic diseases in the United Kingdom, up to 70-80% would need support forself-care (Watkins, 2004).

Obesity, mortality and life expectancy

Unhealthy diets, sedentary lifestyles and obesity are responsible for aconsiderable proportion of the burden of ill health and mortality describedhere. The largest existing study of the link between obesity and mortality,covering close to one million adults in Europe and North America, came to theconclusion that mortality increases steeply with BMI once individuals crossthe 25 kg/m2 threshold (the lower limit of the overweight category)(Prospective Studies Collaboration, 2009). The lifespan of an obese person witha BMI between 30 and 35 is two to four years shorter than that of a person ofnormal weight. The gap increases to eight to ten years for those who areseverely obese (BMI of 40-45), mirroring the loss of life expectancy suffered bysmokers. An overweight person of average height will increase their risk ofdeath by approximately 30% for every 15 additional kilograms of weight.

The link is not as strong beyond age 70 (Stevens et al., 1998; Corrada et al.,2006). Many cross-sectional studies of older individuals have even found alower mortality among the overweight and those who are mildly obese than innormal weight individuals the so-called obesity paradox althoughdetailed longitudinal studies have shown that this is mostly an effect of theweight loss associated with chronic diseases (Strandberget al., 2009).

The overall impact of the obesity epidemic on trends in life expectancy isstill somewhat uncertain, despite the large amount of evidence gathered in


30/270



recent years. A widely cited analysis published in a leading medical journalpredicted that the rise in obesity will lead life expectancy to level off or evendecline during the first half of this century in the United States (Olshanskyet al., 2005). Roughly at the same time, the UK Department of Health claimed

that if the growth of obesity continued unchanged, projected increases in lifeexpectancy to 2050 would have to be revised downwards by over five years(UK Department of Health, 2004). More recent estimates, however, are not sopessimistic. A detailed model-based analysis for England concluded that theloss of life expectancy due to increasing obesity will more likely be in the orderof a fraction of a year by 2050 (Foresight, 2007). A US-based analysis estimatedthat the growth of obesity will offset the positive effects of falling smokingrates, but the net effect will be that increases in life expectancy projectedby 2020 will be held back by less than one year (Stewart et al., 2009). Overall,

downward trends in mortality from a range of chronic diseases are likely tocontinue to prevail over the negative effects of the obesity epidemic, althoughit is unquestionable that progress in longevity would be much faster if fewerpeople were overweight.

However, a growing body of research shows that the impact of obesity ondisability is far larger than its impact on mortality (Gregg and Guralnik, 2007).The obese not only live less than their normal weight counterparts, they alsodevelop chronic diseases earlier in life and live longer with those diseases andwith disability (Vita et al., 1998). In ten European countries, the odds ofdisability, defined as a limitation in activities of daily living (ADL), are nearlytwice as large among the obese as in normal weight persons. The odds arethree to four times as large in men and women who are severely obese(Andreyeva et al., 2007). In the United States, the obese did not benefit fromgeneral improvements in cardiovascular health as much as those with normalweight did. While disability decreased in the latter group, it increased amongthe obese between the late 1980s and the early 2000s (Alley and Chang, 2007).At age 70, an average obese person can expect to live over 40% of their residuallife expectancy with diabetes, over 80% with high blood pressure and over 85%with osteoarthritis, while the corresponding shares for a normal weightperson are 17%, 60% and 68% (Lakdawalla et al., 2005).

The economic costs of obesity

The strong association between obesity and chronic diseases suggeststhat the obese are likely to make a disproportionate use of health care, leadingto a substantially larger expenditure relative to normal weight individuals.A wealth of studies has shown this based on data from at least 14 OECD

countries and some non-OECD countries, mostly focusing on medical careexpenditures. However, the question of the economic impact of obesity is notso simple when addressed over the lifetime and at a population level.


31/270



Estimates based on widely different approaches and methods suggestthat obesity is responsible for approximately 1% to 3% of total healthexpenditure in most countries, with the notable exception of the UnitedStates, where several studies estimate that obesity may account for 5% to 10%

of health expenditures (Tsai et al., 2010). At the individual level, an obeseperson incurs health care expenditures at least 25% higher than those of anormal weight person, according to a range of studies from a variety ofcountries (Withrow and Alter, 2010). When production losses are added tohealth care costs, obesity accounts for a fraction of a percentage point of GDPin most countries, and over 1% in the United States. The figure rises to over 4%in China, according to one study of the economic impact of overweight (ratherthan obesity), which estimated production losses in the region of 3.6% of GDP(Popkin et al., 2006; Branca and Kolovou Delonas, forthcoming).

The lifetime perspective

Because of the time lag between the onset of obesity and related healthproblems, the rise in obesity over the past two decades will mean higherhealth care costs in the future. Taking the example of England, the costslinked to overweight and obesity could be as much as 70% higher in 2015relative to 2007 and could be 2.4 times higher in 2025 (Foresight, 2007).

Only a few of the many studies exploring health care costs associated with

obesity have taken a lifetime perspective. These are all model-based studies,and unfortunately their results are not fully consistent, leaving a great deal ofuncertainty on the long-term impacts of obesity. Two studies published in 1999,both based on US data, suggest that obesity increases lifetime expenditures(Thompson et al., 1999; Allison et al., 1999). At least one of these studies (Allisonet al., 1999) accounts for the disease and health care implications of the longerlife expectancy of people who are not obese, reaching the conclusion that afterage 80 the expected health care expenditures of a non-obese person outgrowthose of an obese person, as the gap in mortality between the two increases

with age. However, the health care expenditures incurred by the obese at earlierages are so much greater than those of the non-obese that, on balance, theobese still have higher lifetime costs.

This conclusion is in line with the findings of a later study (Lakdawallaet al., 2005) that entailed a simulation analysis for a cohort of 70-year-oldsbased on data from the US Medicare Current Beneficiary Survey (MCBS). Thestudy concluded, perhaps unsurprisingly, given its focus on individuals whowere still alive at age 70, that an overweight (but not obese) person has healthcare expenditures about 7% higher than those of a normal weight person,

during the course of their remaining life spans, while the expenditures of anobese person are over 20% higher than those of a normal weight person.However, a further study published in 2008, based on data from the


32/270



Netherlands, found that decreased longevity of the obese makes them likelyto incur lower health care expenditures than the non-obese, over a lifetime(van Baal et al., 2008). According to this study, an average obese person, duringtheir entire life span, will incur 13% lower health expenditures than a normal

weight person, but 12% higher than an average smoker. The sign of thesedifferences did not change in the study under a wide range of assumptions.

Cost-of-illness (COI) studies like the ones described here do provide someuseful information, but is this the information policy makers really need todevise sound prevention strategies? When a study claims that obesity isresponsible for a given amount of health care expenditure, or that obesity isassociated with X% higher health care expenditures, what these claims reallymean is the following : If there were a treatment that made all obese peoplenon-obese and equivalent in health to people who had never been obese, andif this treatment cost nothing to apply, and it were given to all obese people,then in the immediately subsequent time period direct health care costswould be reduced by [X%] (Allison et al., 1999). This hypothetical situation, ofcourse, is very different from the reality policy makers face. Any preventionprogramme, at best, will produce a marginal shift in peoples levels of risk. Ifprevention is successful in moving a certain number of people from obesity topre-obesity, or from the latter to normal weight, those who change theircondition are likely to be the ones who used to be borderline above thethreshold, and their change in weight will probably take them just slightlybelow the same threshold. The changes in health care expenditures followinga real preventive intervention are unlikely to bear much of a relationship withthe estimates provided by COIs.

In the work which led to this book, the OECD deliberately avoidedproducing new generic estimates of health care expenditures, or costs,associated with obesity. Rather, it focused on estimating how specific forms ofprevention may potentially modify existing health care needs andexpenditures, as part of a broader economic analysis in which the costs of

prevention are contrasted with its effectiveness. The methods and findings ofthis work are illustrated in Chapter 6.

The implications for social welfare and the role of prevention

OECD health care systems offer a wide range of treatments for chronicdiseases, aimed at minimising their consequences. Many treatments generatebenefits that justify their costs, notably in terms of quality of life. Still, the needto develop ever better ways to improve quality of life must inevitably confrontthe question of resources: are there limits to what can be spent on improving

the quality of life and extending the life expectancy of those who suffer fromchronic diseases? How do investments in prevention fit into the equation?


33/270



Few countries, if any, have similarly organised systems for the preventionof chronic diseases, although many initiatives have been taken to counterspecific risk factors. As the burden of chronic diseases increases, and associetal expectations in terms of quality of life and longevity also increase,

prevention may offer a valuable alternative to treatment, especially since inprinciple, it has the potential for increasing well-being and longevity evenmore than treating existing disease.

However, the costs and benefits associated with prevention are notalways as obvious as many would think. Unlike treatment, prevention doesnot target diseases, but aims at modifying the conditions that make diseasepossible or likely, such as living conditions, lifestyles and the education peoplereceive. Changing these often involves some kind of individual sacrifice.Examples may include switching from motorised transport to walking orcycling; opting for home cooked meals rather than ready-made and fast foodrestaurant meals; walking an extra distance to buy fresh produce which maynot be available in the neighbourhood; and many others.

Health is not everything

The obesity epidemic is at least in part the result of changes that may bepositive in themselves. Food has become more plentiful and food prices havefallen dramatically. Food is produced and delivered in ways which have cut the

time people have to spend preparing meals, at a time when employmentamong women, who have traditionally done and still do most of foodpreparation, has been steadily on the rise. In 1965, a married woman whodidnt work spent over two hours per day cooking and cleaning up from meals.In 1995, the same tasks take less than half the time (Cutler et al., 2003). For anincreasing number of people, labour is no longer a synonym for work, as jobshave become less and less physically demanding. Motorised transport iscommonplace, even to the local grocery store or school. Obesity, to a certainextent, is a side effect of these and other changes, which Philipson and Posner

(2008) call the positive aspects of the growth in obesity. If, hypothetically,those changes were to be reversed for the sake of a slimmer population, on thewhole, people would be worse off.

A central tenet in an economic approach to prevention is the recognitionthat improving health is not the sole, and often not the most important, goalof human life. Individuals wish to engage in activities from which they expectto derive pleasure, satisfaction, or fulfilment, some of which may beconducive to good health, others less or not at all. Health is complementarywith many forms of non-health-related consumption. It is necessary for

individuals to flourish as consumers, parents, workers, and in othercapacities. But activities from which individuals derive pleasure andfulfilment may also be in conflict with health. Some of these are fairly obvious,


34/270



such as smoking, drinking to excess, or indulging in unhealthy eating.Prevention will inevitably affect the pursuit of activities that are in conflictwith health. As a consequence, individuals will be inhibited to some degreefrom enjoying those activities.

The benefits of prevention over time and across social groups

Why should people change their ways of life? What does prevention haveto offer in exchange for the sacrifices it imposes on individuals? The benefitpeople derive from prevention is not an immediately tangible improvement intheir condition. Rather, it is the prospect of a reduced risk of developingcertain diseases sometime in the future. Both the size of the risk reduction,often relatively small, and the time required for such risk reduction tomaterialise, make it difficult for people to fully appreciate the value ofprevention. Peoples attitudes towards risk, and their preferences concerningoutcomes that may occur at different points in time, have a great influence onthe perceived value of prevention.

The impact of prevention on social welfare depends on the balancebetween the costs of prevention, including the sacrifices imposed on thosewhose environments and lifestyles are affected, and the value attached tofuture risk reductions. Good prevention practices are those which provide realopportunities for increasing social welfare, by ensuring the value of

prevention is greater than its cost. This is the first and foremost goal ofprevention. In addition, prevention may provide opportunities for improvingthe distribution of welfare, or some component of it, such as health, acrossindividuals and population groups.

Health disparities are ubiquitous and persistent in OECD countries, andmany governments have made commitments to reducing them on equitygrounds. Prevention always has an impact on the distributional aspects ofhealth and welfare. Different individuals have different probabilities ofdeveloping chronic diseases, and have different health expectancies once

diseases occur. Different individuals also respond differently to preventiveinterventions, and some will gain more than others from prevention. Thesedistributional effects need to be accounted for in assessing the value ofprevention, and they should be an integral part of the motivation fordelivering prevention programmes. Prevention can be an effective way ofpursuing equity in health when interventions are carefully designed toachieve this goal.

What economic analyses can contribute

This book provides an economic perspective on the prevention of chronicdiseases linked to lifestyles and obesity. That perspective is about more than


35/270


36/270



On the other hand, the economic approach taken here recognises thatindividual lifestyles are subject to influences and constraints that may preventpeople from making the choices that would maximise their welfare. Theability of individuals (obese and not obese) to make choices that would

maximise their own welfare is limited. Even if all individuals were perfectlyrational, the environment in which they live could still prevent them frommaking the best possible choices. ODonoghue and Rabin (2003) emphasisethat economists will and should be ignored if [they] continue to insist that itis axiomatic that constantly trading stocks or accumulating consumer debt orbecoming a heroin addict must be optimal for the people doing these thingsmerely because they have chosen to do it. The same applies to obesity. Itcannot be assumed that all those who become obese willingly accept this as anecessary consequence of behaviours from which they otherwise derive

satisfaction and fulfilment.

Markets can fail

Economics interprets peoples choices and interactions with theirenvironment as market dynamics. There are strong indications, and someempirical evidence, as discussed in Chapter 4, that the market mechanismsthrough which individuals make their lifestyle choices (whether or not moneyis involved), may sometimes fail to operate efficiently. Obesity is partly theresult of these failures, interpreted in this book as market failures, potentiallylimiting the ability of individuals to maximise their own welfare.

Information failures provide a good example of what we mean by marketfailures. The assumption that the consumer has adequate informationconcerning the health effects of food and physical activity is not alwaystenable. But even if the information is complete and unambiguous, manyconsumers may not have the tools needed to use the information provided totheir best advantage. For instance, many consumers would find it difficult tosay whether energy dense and high calorie are the same thing. This is not

just a question of lack of education. In a survey of 200 primary-care patients inthe United States, two-thirds of whom had been to college, only 32% couldcorrectly calculate the amount of carbohydrates consumed in a 20-ouncebottle of soda that had 2.5 servings in the bottle. Only 60% could calculate thenumber of carbohydrates consumed if they ate half a bagel when the servingsize was a whole bagel. (Rothman et al., 2006).

The reasons most people gave for these misapprehensions were that theydid not understand the serving size information, they were confused byextraneous material on the label, and they calculated incorrectly. Information

failures may contribute to the adoption of unhealthy behaviours and lifestylesthrough inadequate knowledge or understanding of the long-termconsequences of such behaviours.


37/270



The problem of self-control

Among the many reasons why people ignore sound advice on health andnutrition, even though they are aware of the economic and health costs

involved, lack of what we commonly call self-control is an obvious one. Peoplegenerally prefer an immediate benefit to a delayed one, even if the later one islarger. Likewise, they discount the longer-term negative consequences of anact that procures immediate gratification. Even if people understand thenegative consequences of eating too much or not exercising, this counts lessthan the more immediate pleasure or other benefit they obtain fromconsumption (ODonoghue and Rabin, 1999; Scharff, 2009).

A key characteristic of people who lack self-control is procrastination(Ariely, 2008, Chapter 6). Those who have poor self-control do not lack

knowledge and information, they are often perfectly aware what they are doingor not doing is bad for their health in the long run, and they are willing andready to change their behaviour, in the future. And they truly believe in theircommitment to change. But when tomorrow comes, of course, they are nolonger prepared to change. This inconsistency of preferences over time, whichis the cause of procrastination, is what makes people with poor self-controlespecially vulnerable to the influences of an obesogenic environment.*

The importance of self-control and ability to delay gratification, is wellexemplified by the famous marshmallow experiment (Mischel et al., 1992).

Pre-school children who were able to refrain from eating a marshmallow whenthey were offered one, in order to gain a second marshmallow reward later,grew up with fewer behavioural problems and a better school performancethan children who were not able to delay gratification. Although obesity wasnot among the outcomes directly assessed in the study, the experiment isrelevant to the issue of weight gain because it shows that self-control is animportant feature of personality, linked to long-term behavioural and socialoutcomes, of which obesity is very likely to be one amidst ever increasingenvironmental pressures.

* ODonoghue and Rabin (1999) observe that most behaviours suggesting the presenceof self-control problems might also be explained in a framework of time-consistentpreferences. For instance: suppose a person becomes fat from eating largequantities of potato chips. She may do so because of a harmful self-control problem,or merely because the pleasure from eating potato chips outweighs the costs ofbeing fat. Procrastination, however, is a clear sign of present-biased preferencesand poor self-control. In practice, the existence of present-biased preferences is

overwhelmingly supported by psychological evidence, and strongly accords tocommon sense and conventional wisdom and even relatively mild self-controlproblems can lead to significant welfare losses.


38/270



External costs of obesity

The obese do not pay the full price attached to their condition. Society atlarge picks up the externalities bill. Externalities linked to obesity may result

in the social or other costs and benefits not being fully reflected in their privatecosts and benefits to individual consumers. For example, a fat person needsmore room on public transport than a thin one, but does not pay a higher pricefor the ticket (although some airlines are introducing extra charges for peoplewho do not fit standard seats). This is a negative externality of being fat, as arethe additional costs to health systems of obesity related diseases (or to hospitalsof having to buy equipment to cope with larger patients).

Fiscal externalities are potentially the most important ones. When healthcare is funded through public expenditure, the cost of the additional health

care needed by an obese person is borne by taxpayers. If an insurance plan orother third party payer is involved, the cost will be shared among all thosecovered by the plan, who pay a premium for their care. However, as discussedbefore, it is still unclear whether the additional health care expendituresgenerated by obesity may or may not be offset by decreased expenditures laterin life, due to premature mortality.

Externalities are also associated with the social mechanisms which makeunhealthy behaviours spread within families, social networks and peer groups asa true multiplier effect. These external costs are very difficult to quantify, but no

less important than others which translate more easily into monetary figures.Externalities generally provide a strong justification for considering

interventions. Evidence of important externalities from smoking and alcoholabuse, among other things, has made possible the implementation of severerestrictions on tobacco and alcohol consumption. Virtually all market andrationality failures will translate either into an excessive or a too limitedconsumption of lifestyle commodities such as food and physical activity,relative to the levels that would be socially desirable. Actions aimed atcorrecting the effects of those failures may tackle directly the mechanisms

through which failures manifest themselves, for instance, by providinginformation when this is lacking or by making individuals pay for the negativeexternal effects of their own consumption, possibly through taxation.

However, it is not always possible, or effective, to act directly on thosefailures. Prevention may also tackle failures indirectly, by acting on anyrelevant determinants of health, to redress the initial overconsumption orunderconsumption. For instance, when information is too complex to becommunicated effectively, the effects of poor information on consumptionmay be compensated by using taxes or other financial incentives.


39/270



Identifying the determinants of obesity

An economic approach to obesity and prevention seeks to identify thedeterminants of obesity those which have changed over time, contributing

to the development of the obesity epidemic, as well as those wh

Documents

Obesity and the Economics of Prevention